Parental Eating Disorders and Broad Autism Phenotype Traits: Is There a Link?

Background: Family members of individuals with ASD demonstrate increased social-communicative problems and multiple neuropsychiatric impairments. One class of neuropsychiatric disorders, eating disorders (ED), has not been well-studied in ASD relatives. However, several studies have shown that women with ED, especially anorexia nervosa, display clinically significant scores on measures of ASD symptoms. This study is the first to systematically evaluate the relationship between ED and the broad autism phenotype (BAP) in first-degree relatives.

Objectives: This study will: (1) compare autism trait measures for parents of individuals with ASD who have ED to those without ED; and (2) compare clinical outcomes for the offspring of these two groups.

Methods: Our dataset consisted of 2623 families from the Simons Simplex Collection (SSC), a collection of families containing one individual with ASD and no other first-third degree relatives with an ASD diagnosis.

As part of SSC ascertainment procedures, family histories and multiple measures of ASD traits were collected.  Family history interviews included questions about the presence of ED in parents and their relatives. The Broad Autism Phenotype Questionnaire (BAP-Q) and Social Responsiveness Scale: Adult Research Version (SRS:ARV) measured ASD traits.

Descriptive analyses compared ED and non-ED groups on BAP-Q total and subscale scores (Aloof, Pragmatic Language, and Rigidity), and SRS total and subscale scores (Awareness, Motivations, Mannerisms, Cognition, and Communication). The proportion of individuals with clinically significant BAP-Q scores in both groups was also compared. Finally, clinical outcomes in offspring as a function of parental ED were examined.

Results: In total (N=2623), no fathers and 57 (2.2%) mothers had a positive history of ED. Comparison of BAP-Q scores showed that mothers with ED had significantly higher BAP-Q Total (p=6.6^10-5) and subdomain (Aloof p=0.003; Pragmatic p=1.4^10-4; Rigid p=0.013) scores. In contrast, mean SRS total and subdomain scores did not differ between the ED and non-ED mothers.

Among mothers with an ED, there was a significantly higher proportion with an elevated BAP-Q Total score (and elevated subdomain scores). The most significant differences were noted for the BAP-Q Total score [22.8% of those with an ED were positive for the BAP versus only 8.8% without an ED (p=0.0003)] and the BAP-Q Rigid subdomain score [21.1% of those with an ED were positive for this trait vs. 5.6% without an ED (p=1^10-6)]. Comparisons of offspring of mothers with ED vs. those without ED showed no significant differences in sex, adaptive functioning, or behavior problems.  

Conclusions: Our results show that ED were restricted to mothers and lower than population estimates. For mothers with an ED, BAP-Q scores were significantly higher than mothers without an ED and were more likely to meet criteria for the presence of the BAP trait. By contrast, the groups did not differ on any SRS variables.

These results converge with findings of increased ASD features among women with ED. The overlap of ED and elevated BAP-Q scores merits further study to determine if there is also biologic overlap (i.e., common genes or pathways) and which BAP features may heighten risk for ED in select individuals.